|
The collaboration between Oracle and CERN is a particularly
long-standing and fruitful one. The partnership started 26 years
ago and Oracle decided to join CERN openlab from its early
beginnings, in 2003. Oracle is now about to enter the third
phase of the programme starting in January 2009, writing a new
chapter to the openlab story.
The CERN-Oracle partnership in the
openlab proved to be an incubator for innovation. Examples of
this are the Maximum Availability Architecture technologies
(Real Application Clusters – RAC, Streams and Data Guard) which
are now used in production on a worldwide scale for key elements
of data production and processing for the
Worldwide LHC Computing Grid (WLCG).
The WLCG is a global collaboration
that connects and combines the IT power of more than 140
computer centres in 33 countries ; these computer centres are
structured in a tiered system, with the CERN computer centre as
the tier 0 centre of the grid. Its mission is to build and
maintain a data storage and analysis infrastructure for the high
energy physics community that will use the
Large Hadron Collider (LHC)
at CERN.
Oracle Streams is the main
replication technology used for relational data distribution in
the framework of the LHC. Different kinds of data (vital for
data processing and analysis) such as detector conditions,
calibration data, alignment information and accelerator
conditions are stored in Oracle RAC databases and are replicated
through a distributed database infrastructure between CERN and
ten of the eleven tier 1 computer centres all over the world.
Furthermore, the openlab collaboration enabled CERN to standardise and greatly improve the monitoring
of the Oracle infrastructure with the Enterprise Manager 10g
Grid Control product, resulting in less time spent firefighting
and more time adding value for the organisation. Oracle
has helped through several major software upgrades and the
migration from Solaris to Linux, as well as giving advice on
extending the products’ capabilities. In return, the teams at
CERN gather a lot of real user experience and shared this
knowledge at conferences and with the Oracle product and
development teams. This has guided product enhancements, and
helped Oracle to find new solutions, which they can pass on to
other customers.
Within the Joint Software Testing programme, where Oracle provides early
versions of their software, CERN have been able to confirm
compatibility of some of its most critical applications with
Oracle’s newest functionality. In particular, testing Oracle’s
storage “offload operation” system - Exadata
– meant
the potential benefits could be quantified for some of the most
data-insertion intensive applications CERN relies on. Thanks to
this testing programme, the teams have provided valuable
feedback on the early versions of this new Oracle technology,
which was included in the software before the product was
released. Since the Oracle VM announcement last year, the teams
involved in CERN openlab have been working extensively towards
virtualization. A first implementation for some of CERN’s
non-critical databases will result in a reduction of hardware
and infrastructure costs.
From 2009 onwards, with Oracle
entering CERN openlab III, further developments are expected in
these domains as well as studies related to the upcoming Oracle
11g release 2.
To fulfill the ambitious objectives of each CERN openlab
phase, innovation and motivation have been key factors and
Oracle has been very forthcoming about funding highly proficient
and proactive fellows. The CERN openlab fellows provided the
bridge between CERN and Oracle's own R&D division during openlab
I and extended and deepened the relationship between the two
organizations during the second phase. In the third phase of the programme,
Oracle will continue to fund fellows and work closely with the
two CERN IT groups involved in the previous phases,
the Data Management Group
and
the Database & Engineering Services
Group.
|